DM-37387: A tester script using LATISS data in /repo/embargo

[hsinfang]

A first part of this ticket to get test data into s3://rubin:rubin-pp-users/central_repo

[hsinfang]

An obvious problem here is that some export scripts are somewhat similar but I haven't bothered to consolidate them. In particular this new script and make_hsc_rc2_export.py are quite close and I probably should combine them to one.

[kfindeisen]

The initial script looks very clean and clear. However, I don't think I understand the --target-repo argument or the code that depends on it. Could you please clarify how it works, and when the caller might opt not to use it?

[hsinfang]

This script makes an export.yaml file and this export file may be used to populate an empty freshly created butler repo, or to add more datasets into an existing repo. The motivation I added the --target-repo argument was for the second case, and I wanted to skip the datasets that already exist in the target repo (s3://rubin:rubin-pp-users/central_repo to be precise). It turns out that the only existing datasets are refcats for what I'm doing in this ticket. That's expected; the script as-it was used to verify that.
I think I'm struggling a bit between making the script more generic versus for the very specific usage in mind. The script assumes a lot about the data already. So maybe checking every dataset doesn't make sense.

[hsinfang]

@kfindeisen may you please take another look? Thanks to your feedbacks I think the script looks better now. It still assumes some peculiarities of /repo/embargo for those datasets and collections, but probably the script could be made useful relatively easily when we want to do something similar in the future.

[kfindeisen]

Looks good, thanks! Minor comments only.